Qwen

Qwen3 235B A22B

Qwen3 235B A22B is an open-weights 235B-parameter model by Qwen. It is listed as in-progress on the local.ai leaderboard.

MoEopen weights
Parameters
235B
MoE · 22B active parameters
Benchmarks (local.ai snapshot)
τ²-bench
GAIA
GDPval
HuggingFace metadata · updated 2025-07-26
repo
Qwen/Qwen3-235B-A22B
license
apache-2.0
downloads
525,827
likes
1,108

Metadata refreshed from the HuggingFace API (snapshot 2026-08-15T15:53:42.405Z). Context length and long-form descriptions are the next pipeline stage.

Does it fit your hardware?

Detect or select your GPU first. Green = weights + context headroom, yellow = tight, grey = does not fit.

Q4_K_M 136.3GQ6_K 211.5GQ8_0 258.5GFP16 470G
Quantization≈ File sizeNote
IQ2_XXS63.5 GBExtreme compression, largest quality loss
IQ3_XXS84.6 GBVery small
Q3_K_M108.1 GBSmall
NVFP4117.5 GBNVIDIA 4-bit
Q4_K_S126.9 GBCommon
Q4_K_M136.3 GBMost common sweet spot
Q5_K_M178.6 GBBalanced
Q6_K211.5 GBNear-lossless
Q8_0258.5 GBNear-lossless
FP16470 GBOriginal weights

Actual files on HuggingFace

FileSize
model-00030-of-00118.safetensors3.72 GB
model-00111-of-00118.safetensors3.72 GB
model-00029-of-00118.safetensors3.72 GB
model-00110-of-00118.safetensors3.72 GB
model-00034-of-00118.safetensors3.72 GB
model-00039-of-00118.safetensors3.72 GB
GGUF: unsloth/Qwen3-235B-A22B-GGUF · 145,802 downloads
FileSize
BF16/Qwen3-235B-A22B-BF16-00001-of-00010.gguf46.15 GBdownload
BF16/Qwen3-235B-A22B-BF16-00002-of-00010.gguf46.21 GBdownload
BF16/Qwen3-235B-A22B-BF16-00003-of-00010.gguf46.35 GBdownload
BF16/Qwen3-235B-A22B-BF16-00004-of-00010.gguf46.35 GBdownload
BF16/Qwen3-235B-A22B-BF16-00005-of-00010.gguf46.35 GBdownload
BF16/Qwen3-235B-A22B-BF16-00006-of-00010.gguf46.35 GBdownload
BF16/Qwen3-235B-A22B-BF16-00007-of-00010.gguf46.35 GBdownload
BF16/Qwen3-235B-A22B-BF16-00008-of-00010.gguf46.35 GBdownload
BF16/Qwen3-235B-A22B-BF16-00009-of-00010.gguf46.35 GBdownload
BF16/Qwen3-235B-A22B-BF16-00010-of-00010.gguf21.20 GBdownload
IQ4_XS/Qwen3-235B-A22B-IQ4_XS-00001-of-00003.gguf46.55 GBdownload
IQ4_XS/Qwen3-235B-A22B-IQ4_XS-00002-of-00003.gguf46.46 GBdownload
GGUF: unsloth/Qwen3-VL-235B-A22B-Instruct-GGUF · 14,087 downloads
FileSize
BF16/Qwen3-VL-235B-A22B-Instruct-BF16-00001-of-00010.gguf46.15 GBdownload
BF16/Qwen3-VL-235B-A22B-Instruct-BF16-00002-of-00010.gguf46.21 GBdownload
BF16/Qwen3-VL-235B-A22B-Instruct-BF16-00003-of-00010.gguf46.35 GBdownload
BF16/Qwen3-VL-235B-A22B-Instruct-BF16-00004-of-00010.gguf46.35 GBdownload
BF16/Qwen3-VL-235B-A22B-Instruct-BF16-00005-of-00010.gguf46.35 GBdownload
BF16/Qwen3-VL-235B-A22B-Instruct-BF16-00006-of-00010.gguf46.35 GBdownload
BF16/Qwen3-VL-235B-A22B-Instruct-BF16-00007-of-00010.gguf46.35 GBdownload
BF16/Qwen3-VL-235B-A22B-Instruct-BF16-00008-of-00010.gguf46.35 GBdownload
BF16/Qwen3-VL-235B-A22B-Instruct-BF16-00009-of-00010.gguf46.35 GBdownload
BF16/Qwen3-VL-235B-A22B-Instruct-BF16-00010-of-00010.gguf21.20 GBdownload
IQ4_NL/Qwen3-VL-235B-A22B-Instruct-IQ4_NL-00001-of-00003.gguf46.53 GBdownload
IQ4_NL/Qwen3-VL-235B-A22B-Instruct-IQ4_NL-00002-of-00003.gguf46.53 GBdownload

What you save self-hosting this model

Uses the model's own API price when it exists; otherwise the closest commercial equivalent.

Waiting for hardware detection…

RunLocal presents third-party leaderboard data (local.ai / Exo Labs, 2026-08) with its own fit and cost estimates. Verify pricing, licenses and file sizes at the source before deploying a model.